Category Archives: Thoughts on the World

Are We Becoming the Eloi?

Hunter S Robot writing blog

Are we becoming the feckless race from The Time Machine? Can we resist the temptation?

Background

This article is the result of convergence of two separate prompts. About a week ago I was bemoaning our increasing dependency on automation and our general inability to take charge and either fix or work around broken things, and I thought of the Eloi. In H G Well’s The Time Machine, published in 1895, he imagined a distant future human race who live an apparently idyllic existence, but who have lost any ability to help themselves, and are to some extent just farm animals for the Morlocks, who prey upon them. In some ways, we may be approaching the society Wells foresaw, but only 130 years, rather than 800,000 years later. In my writing ideas list, I wrote down “Are We Becoming the Eloi?”

The second prompt was the recent publication of Matt Shumer’s essay Something Big is Happening in AI. That asserts that we are rapidly reaching (or may have already reached) “the singularity”, where AI starts to evolve at a speed outstripping our control, and displaces a large proportion of what we used to call “knowledge workers”, with dramatic societal impacts.

I was going to write the Eloi article myself, honest I was. But having read Matt’s article, I couldn’t help thinking “could AI do this?”. Could it do the research (I haven’t read The Time Machine for at least 40 years), reach reasoned conclusions with which I agree, and write in a style reasonably mimicking my own?

So I fired up the free version of ChatGPT with the following prompt:

Write an essay of 1300-1400 words, on the subject “Are We Becoming the Eloi?” Take as reference the race of the Eloi in H G Wells’ The Time Machine. Compare current societal trends such as an increasing reliance on automation, increased distance between users and underlying technology (with a commensurate inability to act if the technology is not working) and the increased anxiety and mental health issues of younger generations. Write in a style similar to this article: https://www.andrewj.com/blog/2025/adjustable-architecture-and-minimising-waste/

The result (see here) is very impressive, a detailed, reasoned essay analysing ways in which we are starting to resemble the Eloi, but also identifying key differences and ways in which we might escape that fate. It’s quite academic, with some philosophy references beyond my knowledge, but at a stretch I might pass it off as “all my own work”.

The main problem is stylistic. The article is more academic than I’d write, complete with valuable references. It’s wordy, in need of at least as much editing for verbosity and repetition as my own first drafts do. There are also odd issues with terminology. The article repeatedly uses the word “agency” with a meaning I don’t usually adopt, and has also invented a clumsy term which I would never use: “Eloi-ness”.

However, there’s a simple fix. In Red Dwarf there’s a wonderful scene where the rest of the crew are faced with two Dave Listers – one the real thing, the other a dangerous simulant who can copy his appearance and personality. Both are holding guitars. Both play, one badly and the other is sublime. Without a word the crew unanimously shoot the latter. The real Dave Lister can’t play the guitar, but he thinks he can, so that’s what the simulant does.

So I just need to instruct ChatGPT to write the article not as I actually write, but as I like to think I write… Prompt 2:

Write a similar essay, but make it less academic, and write in the style of Hunter S Thompson.

To put it bluntly, that nails it. The essay, below, captures almost all the concepts I wanted to incorporate, in a style which is both blunter and smarter than my own, but not so far removed you might not believe I wrote it myself on a good day.

The essay is presented exactly as generated, excluding minor tweaks to punctuation, and changing a couple of American spellings.

The Essay: Are We Becoming the Eloi?

There is a particular kind of silence that settles over a room when the Wifi dies. Not the peaceful silence of a countryside evening, but the twitchy, electric quiet of people suddenly cut loose from the machinery that explains the world to them. Screens freeze. Conversations stall. Someone begins stabbing at a router like it personally insulted their family. It is in moments like these – tiny technological blackouts – that you can almost hear the faint laughter of H G Wells echoing across the century, because he saw something like this coming.

In The Time Machine, Wells imagined the Eloi: pale, gentle, useless creatures drifting through a pastoral future where every need was met and every challenge long forgotten. They did not work. They did not invent. They did not even wonder very much. They simply lived in ornamental comfort while the unseen machinery of their survival ran elsewhere, managed by another species they barely understood.

It was a grotesque fable about class, complacency, and the slow erosion of capability. For a long time, it felt comfortably fictional: Victorian nightmare fuel, a warning directed at a distant future we were sure we would never resemble.

And yet here we are, living in a world where a great many people cannot boil an egg without consulting a search engine, cannot navigate a city without a glowing blue dot telling them where to turn, and cannot repair even the simplest device because every tool they own is sealed shut with proprietary screws and legal disclaimers. We are not helpless, not yet – but we are drifting in that direction with the serene confidence of people who believe the escalator will run forever.

Automation was supposed to free us. That was the promise: machines would do the dull work, algorithms would handle the complexity, and humans would finally be liberated to pursue creativity, leisure, and higher thought. In theory, it sounds magnificent. In practice, something stranger has happened. Instead of freeing us to become more capable, automation often encourages us to become less involved. It doesn’t just do the work – it removes the need to understand the work at all.

Consider how many systems you rely on daily that you could not repair, replicate, or even explain in broad terms. Your phone connects to a network you don’t understand, which runs through infrastructure you never see, governed by software you never touch, administered by people you will never meet. If the entire structure hiccups, your response is not to diagnose it but to wait – helplessly, irritably – for someone else to fix it. You are a passenger in a machine that does not require your participation, only your subscription.

The Eloi, in Wells’ nightmare future, were not stupid because they were born that way. They were stupid because their environment stopped demanding anything from them. Muscles that are never used fade; instincts that are never exercised disappear; curiosity that is never necessary grows quiet. The human brain is no exception. When every friction is removed from life – when navigation, memory, calculation, communication, and even entertainment arrive automatically – the temptation is not to grow wiser with our extra time. The temptation is to drift.

Look at the modern office worker staring blankly at a spinning “loading” icon, as if their entire professional identity has been unplugged along with the server. Watch the commuter pacing the platform because the ticketing system crashed and no one remembers how to process anything manually. Observe the mild panic that sweeps through a café when the contactless payment terminal fails and suddenly the ancient technology known as “cash” must be summoned from dusty wallets. These are not catastrophic events. They are small inconveniences. But the reaction they provoke, the sense that the world has temporarily become unmanageable, tells a deeper story about how far we have stepped back from the machinery that runs our lives.

Meanwhile, younger generations grow up inside systems that are even more abstracted. They do not learn how things work; they learn how interfaces behave. They swipe before they read. They troubleshoot by restarting the app. They store their memories in cloud servers whose physical locations might as well be on Mars. This is not a moral failing; it is simply the environment they were handed. But environments shape people, and an environment that constantly cushions failure also quietly erodes resilience.

There is a peculiar irony here. Never in history have humans possessed more knowledge, yet never has the average individual needed to know less to function day-to-day. A person in the nineteenth century might have needed practical familiarity with tools, navigation, mechanical systems, and local infrastructure just to survive ordinary life. Today, survival depends more on passwords than on skills. Forget the password, and you are locked out of your own existence. Forget how to mend a device, and you simply replace it. Forget how to navigate without GPS, and you circle the same block until the satellite signal returns like divine revelation.

This creeping dependency does something subtle to the psyche. When you no longer feel capable of solving problems directly, when solutions always arrive from outside, delivered by software updates or distant technicians, you begin to feel smaller in relation to the world. Problems seem larger. Systems feel opaque. Control appears theoretical rather than practical. Anxiety flourishes in precisely this kind of psychological climate: a world that is enormously powerful, endlessly complex, and largely inaccessible to individual intervention.

The rising tide of anxiety among younger people is often explained in terms of social media pressure, economic uncertainty, or educational stress. All of those factors matter. But there is another quieter ingredient: the persistent feeling of operating inside systems you cannot influence. When everything depends on infrastructure you cannot see and algorithms you cannot challenge, the world begins to feel less like something you shape and more like something that happens to you. The Eloi lived like that – pleasantly, superficially happy, but fundamentally detached from the forces that determined their fate.

Of course, the comparison has limits. We are not lounging in marble ruins eating fruit while subterranean creatures manage civilization. We build extraordinary things. We invent, argue, protest, experiment, and obsessively redesign our own technologies. But cultural direction matters, and the direction of modern convenience is unmistakable: fewer manual skills, less mechanical understanding, more automated decision-making, and an ever-widening gap between what systems do and what users comprehend.

The most dangerous aspect of this trend is not that machines are getting smarter. It is that humans are becoming more comfortable not needing to be. Intelligence, like strength, fades when unused. If every task that once required patience, memory, or ingenuity is replaced by a single button labelled “Optimize,” those faculties don’t vanish overnight. They simply stop developing. Over time, a society can become incredibly sophisticated at the macro level while individuals become strangely passive at the micro level, surrounded by miracles they cannot recreate and comforts they cannot maintain.

And yet, the situation is not hopeless. Far from it. The key difference between us and Wells’ fictional future is awareness. The Eloi did not know what they had lost. We still do. We still teach engineering, repair, science, navigation, and craftsmanship. We still debate the ethics of automation and the limits of artificial intelligence. We still celebrate people who understand how things work, even if the average citizen doesn’t always feel compelled to learn those skills personally. Awareness is the thin line that keeps a technologically advanced civilization from sliding into ornamental dependency.

The question, then, is not whether technology will continue to automate life – it absolutely will – but whether we choose to remain mentally and practically engaged with the systems we create. Convenience is not the enemy; indifference is. The danger is not that machines help us, but that we gradually forget how to help ourselves.

Picture the world fifty years from now. It could be a dazzling landscape of automated infrastructure, personalized AI systems, and seamless logistics, with citizens who understand enough of the underlying mechanics to intervene, repair, and innovate when necessary. Or it could be a comfortable garden full of passive consumers, bewildered whenever the machinery pauses, waiting patiently for unseen specialists to restore normality. Both futures are technologically advanced. Only one avoids the fate of the Eloi.

The warning hidden in Wells’ strange little novel is not that progress destroys humanity, but that comfort can quietly dissolve the habits that make humanity resilient. Skills unused fade. Curiosity neglected dulls. Agency surrendered becomes difficult to reclaim. And the transformation does not arrive with dramatic fanfare: it happens slowly, politely, wrapped in the reassuring language of efficiency and convenience.

So the next time the Wifi drops and the room falls into that anxious, humming silence, it might be worth resisting the instinctive panic. Look around. Ask yourself how much of the world you can still operate without the invisible scaffolding of automated systems. Not because you intend to reject technology – only a lunatic would do that – but because the ability to function without total dependence is the thin line between mastery and decoration.

The Eloi did not notice the moment they became helpless. That is precisely why we should.

Assessment

The experiment provides further evidence, if it were needed, supporting Matt Shumer’s assertion. Using just two simple prompts the free version of ChatGPT has generated work which I would be proud to claim as my own, honesty permitting.

Much of my professional career hinged upon my ability to consolidate disparate sets of information, analyse and summarise them into a coherent whole, and then clearly communicate the expected impacts and options for action. A large part of that work can now be laid off to an AI.

That said, at least for now there’s a difference between “implement this well-formed idea” and “help me develop this idea”. It’s not clear how AI would fit into my 40-year-old tried and tested analysis process. I start with a blank document, or maybe a vague outline, and dump all sorts of ideas, notes, research material and other people’s contributions at the bottom. I then work down through all the material, delivering a curated and edited version into the document above. I eliminate the irrelevant or duplicated. Some material is included only to be refuted. The text is constantly refactored to ensure flow and coherent arguments. That process in turn generates ideas for further research, tests and writing.

For example, after reading the AI generated essay, I thought of two more potentially rich seams of analysis:

  1. Wells wrote The Time Machine in 1895. Phones and electricity services were a recently-invented novelty. Distant communication was by letter or telegraph. Wells set the later scenes 800,000 years in the future, and failed to predict the exponential, accelerating changes which mean we are confronting these challenges after a mere 130 years. Does the speed of change impact on the effect it will have?
  2. In The Time Machine, the Eloi are supported by the dark, subterranean Morlocks. They manage the hidden technology but also prey upon the Eloi, for food. In a very literal sense the Eloi are not the customers, they are the product. This is clearly a powerful, if extreme, metaphor for the capitalist providers of our tech. That metaphor could be explored further.

AI could clearly be instructed to extend the essay with these concepts, but it didn’t initiate them. For now, I still have to have the ideas.

However, I can imagine an AI which keeps track of my writing ideas, and regularly prompts me with “what do you want to write next?”, takes my notes and comes up with a first draft. It would be trained to mimic my own style, not one to which I aspire. It could automatically prompt the image generator to generate some sample hero images with my regular cast of characters, like the puzzled bear. It could automate the posting process. I could become one of the most prolific bloggers, but at what point is the blog no longer mine, but a computer’s?

There’s an interesting difference between the two versions of the essay. The first, academic, version (see here) goes into quite a lot of detail on the mental health dimension, complete with supporting references. Despite being an explicit element of the prompt, this is only lightly touched upon in the second version. I’ve seen this before: stylistic guidelines can change not only the writing style, but may also impact on tone, direction and content.

The temptation to use AI is potentially overwhelming. Two 1400 word essays were each generated for me in less than a minute. Either could be used without further work. By comparison my introduction and assessment have taken at least 6 hours to write. If I was getting paid for this the temptation would be hard to resist.

Even where I need to write the words myself, I will now use AI for supporting tasks. My blog’s “hero” images are either my own photos, or AI-generated cartoons. If I was writing my project management book today I wouldn’t employ a cartoonist. And that’s another job gone…

However, I’m not sure there’s any evidence yet of “the singularity”. The main reason is I think we’ve got the wrong idea about it. AI doesn’t have to be fully sentient to cause harm. It doesn’t have to launch nukes or push us out of an airlock to cause profound societal change.

Instead, I think we are going to experience a series of inflection points in different disciplines. We may not notice many of them as AI rapidly accelerates towards and passes them, we will only recognise them in the rear-view mirror.

We have to understand how AI will affect society, and that brings us back directly to my own question. Are we becoming the Eloi?

The answer is probably “not exactly”. I don’t think we need fear a future in which we are all completely degenerate, farmed by a successor species, aliens (or machines, as in The Matrix).

But we may be headed very rapidly for a world in which we all display some characteristics of the Eloi, some of the time. We are already in a society in which a substantial subset feel things are done to them, not by them. Some of us will also be Morlocks, unless we somehow arrest the rise of exploitative corporate giants.

The speed of change is significant. Wells imagined a gap of 800,000 years, in which the Eloi have degenerated physically as well as mentally. In reality in that period a better-equipped species of Homo might evolve. Instead, we are trying to manage massive and accelerating mental and societal change with much the same physical provision as the first of our species.

I think the ChatGPT essay sums it up well: “The Eloi did not notice the moment they became helpless. That is precisely why we should.”

We need to be alert to the risks. As individuals we need to try and understand what we can, avoid always taking the easy way out, and sometimes deliberately “do it the hard way”. We also need to make sure those attitudes are inculcated into younger generations. We need to make sure our political decision-makers are also alert to the risks, and not blind-sided by the siren calls of big tech.

We don’t have to become the Eloi. But there’s a very real risk that we will.

View featured image in Album
Posted in Thoughts on the World | Leave a comment

Are We Becoming the Eloi? The Academic Version

Academic robot writing blog

Introduction

This is an appendix to a more complete article, see https://www.andrewj.com/blog/2026/are-we-becoming-the-eloi-the-academic-version/

I was recently bemoaning our increasing dependency on automation and our general inability to take charge and either fix or work around broken things, and I thought of the Eloi. In H G Well’s The Time Machine, published in 1895, he imagined a distant future human race who live an apparently idyllic existence, but who have lost any ability to help themselves, and are to some extent just farm animals for the Morlocks, who prey upon them. In some ways, we may be approaching the society Wells foresaw. At the same time I read another essay asserting that AI is ready to displace a large proportion of what we used to call “knowledge workers”. I decided to test the latter assertion by asking ChatGPT to draft an essay on “Are We Becoming the Eloi?”.

I fired up the free version of ChatGPT with the following prompt:

Write an essay of 1300-1400 words, on the subject “Are We Becoming the Eloi?” Take as reference the race of the Eloi in H G Wells’ The Time Machine. Compare current societal trends such as an increasing reliance on automation, increased distance between users and underlying technology (with a commensurate inability to act if the technology is not working) and the increased anxiety and mental health issues of younger generations. Write in a style similar to this article: https://www.andrewj.com/blog/2025/adjustable-architecture-and-minimising-waste/

The essay below is what resulted. It’s presented exactly as generated.

For various reasons, I then went on to generate a second version, with different stylistic guidelines, which I use in the full article. However some readers may want to compare the two versions of the essay, so I’m providing the first draft here.

Are We Becoming the Eloi?

In H.G. Wells’ The Time Machine, the Eloi occupy a distant future where humanity has, ostensibly, solved all material hardship. They laze in sunlit architecture, eat fruit, play, and live without work — carefree in their quotidian existence but frighteningly hollow in their capacities for curiosity and resilience. Over thousands of years of ease, they have devolved into a species that can barely fend for itself, dependent on the subterranean Morlocks for food, clothing, shelter, and even safety. Wells’ grotesque future serves as a caution against complacency, inequality, and the seductive illusions of technological progress. The question now is less about speculative evolution and more about whether we, in the twenty-first century, are sliding down that same evolutionary slope — into a kind of real-world Eloi-ness.

The Eloi and the Modern Soul

Wells’ Eloi are not merely idle; they lack intellectual curiosity, physical strength, and any meaningful connection to the systems that sustain them. They are the elegant unwashed, unaware that their comfort is the product of unseen labour — the Morlocks’. In The Time Machine, this absence of challenge corrodes their faculties, leaving them childlike and helpless. Their bodies, their minds, and their instincts have atrophied in the absence of need or struggle. This is a chilling image: not merely of laziness, but of degeneration in the philosophical sense — of potential atrophy where capability once lived. (Wikipedia)

At first glance, the parallels to twenty-first-century societies might seem hyperbolic. After all, few of us lack agency outright, and we remain painfully aware of the world’s injustices. Yet, when one looks closely at the interplay between technology and human behaviour today, the metaphor resonates with surprising precision.

Automation and the Loss of Agency

One of the prominent trends of our age is the exponential growth of automation. Machines and algorithms now perform tasks that once defined the substance of human labour: driving, diagnosing diseases, managing finances, even composing music. What was once the domain of specialised tradespeople or rigorous study is now routine output from digital systems. This explosion of automation brings efficiency and convenience — but also an attendant loss of direct engagement with the processes themselves.

Consider the average smartphone user: for most tasks, we never see the mechanics beneath the interface. We swipe, tap, ask an AI, and receive a result. If a system fails, many users are helpless — unable to perform even basic tasks without the intermediary of a screen or app. The analogy to the Eloi begins to crystallise: a population adept at consuming results but alienated from the meanings and processes that produce them. (Mental Health America)

This distance from the underlying systems is not merely a matter of convenience, but of cognitive consequence. The sociologist David Collingridge articulated a dilemma in technology governance: by the time the impact of a technology is evident, it has become so embedded that changing or controlling it becomes difficult. In effect, we trade agency for ease, and the psychological fallout of this trade — the sense that we are passengers rather than drivers — cannot be ignored. (Wikipedia)

Dependency and the Tech Ecosystem

This loss of agency is mirrored in how younger generations relate to technology. Studies from mental health research show that heavy reliance on digital devices and platforms is nearly ubiquitous among young people near adulthood, with a majority reporting a heavy dependence on technology for everyday functions — from social connection to information to entertainment. (Mental Health America)

Some research suggests this dependency is correlated with higher anxiety and emotional distress — not because technology causes mental illness in a simple, direct way, but because compulsive use, constant digital feedback loops, and the loss of control over habits and attention can amplify feelings of anxiety and instability. (Mental Health America)

Most strikingly, modern studies on adolescent screen use link addictive patterns of engagement — not merely hours spent but patterns of compulsive use — to emotional problems, suicidal ideation, and symptoms of anxiety and depression. These findings indicate not only psychological stress but an erosion of resilience and self-regulatory capabilities among youth deeply embedded in digital ecosystems. (The Guardian)

If Wells’ Eloi succumbed to ease and lost their capacities over generations, modern society — especially among young people — is showing hints of that same pattern in behavioural trends: from diminished abilities to persist without external stimuli, to emotional fragility when devices are absent, to a reliance on automated systems rather than independent problem-solving.

The Illusion of Comfort

But here the analogy must be sharpened rather than stretched: we have not (yet) become feeble, incapable hominids. We have, however, cultivated layers of comfort that insulate us from hardship in ways that could subtly reshape character and capability. For the Eloi, every challenge had been abolished by generations of ease; for us, many challenges have been outsourced to technologies that mediate nearly every aspect of life.

As digital assistants manage our calendars, AI tools draft our text, automated systems handle logistics, and apps adapt our environments, the experience of mastering the world recedes. We may gain time in one sphere, but we lose depth in another. Without necessity, struggle becomes optional — and without struggle, the psychological mechanisms that build resilience, critical thought, and emotional strength are underused.

A telling metaphor is our response to technological failure. When our internet goes down, many of us feel not irritation — but panic. A glitch in the GPS can leave us disoriented, not because we lack curiosity, but because we have surrendered our navigational instincts to silicon. This isn’t just convenience — it’s a surrender of agency. (vida.id)

The Mental Landscape of a Digitally Infused Generation

Psychological research suggests that this isn’t an entirely imagined concern. Longitudinal studies show increases in anxiety and depressive symptoms among adolescents in recent years, often concurrent with heavy digital device use and the pressures of online social ecosystems. (Mental Health America)

These associations aren’t simple causal chains — technology doesn’t create mental illness — but the context of addiction-like patterns, comparison-driven social media, and feedback loops tailored for maximum engagement has reshaped the emotional environment in which young people develop. In some respects, what we’re witnessing is not the loss of ability, but the erosion of confidence, presence, and internal grounding: psychological analogues to the Eloi’s loss of survival muscle.

Agency as Antidote

Yet the comparison to the Eloi also reminds us that sheer ease alone does not determine destiny; social context, economic structures, and deliberate choice matter enormously. Wells’ Eloi were isolated on the surface, dependent without awareness of their dependence. By contrast, we remain deeply aware of both the benefits and risks of our technologies.

We debate automation ethics, AI governance, digital addiction, and the dangers of data concentration. We ask whether curricula should teach resilience alongside computational thinking. The very existence of these debates suggests that we are not passive participants in our own evolution — a distinction the Eloi never had. In fiction, the Eloi don’t know what they don’t know. In reality, we are increasingly conscious of our dependencies, anxieties, and the psychological costs of technological integration.

Moreover, human agency is more than the ability to perform tasks. It is the capacity for reflection, for choosing how and why we use tools — and it’s this capacity that remains at the forefront of the discourse around technology, education, and societal development.

Conclusion: A Call to Avoid Eloi-ness

To be an Eloi is not simply to be comfortable; it is to be unaware of the foundations that sustain comfort. That unawareness is the true loss Wells warns against. In many ways, our contemporary moment is not yet that dystopia. We still think, we still worry, we still debate our tools and our futures.

But there is a risk that, without intentional engagement, that awareness could be lost. If we outsource too much of the work of thinking, acting, and navigating life to machines and algorithms, the subtle atrophy of resilience and self-direction could accelerate. Becoming the Eloi would not happen overnight — it is a generational drift, a quiet disengagement from the challenges that shape thought and character.

If we are to avoid that fate, the antidote lies not in rejecting technology, but in reclaiming agency: teaching skills not just of operation but of understanding; cultivating resilience alongside intelligence; and ensuring that the comforts of automation do not become the cocoon that dulls our capacities rather than enabling our potential.

View featured image in Album
Posted in Thoughts on the World | Leave a comment

Evolution of an Image

Central image reprocessed with 2026 software
Camera: Canon EOS 40D | Lens: EF-S17-85mm f/4-5.6 IS USM | Date: 28-04-2008 17:09 | Resolution: 2491 x 3737 | ISO: 200 | Exp. bias: 0 EV | Exp. Time: 1/40s | Aperture: 11.0 | Focal Length: 17.0mm (~27.6mm) | Lens: Canon EF-S 17-85mm f4-5.6 IS USM

My photography mentor, Bob Kiss, recently posted an image of his, taken in Tuscany, of a Tuscan field scene shot through a window, with the light carefully balanced so that you can clearly see both the exterior, and the interior around the window. It immediately reminded me of a not dissimilar shot I took many years ago, at the Fish Pot in Barbados.

While Bob’s shot is relatively recent, and took advantage of modern sensor technology and computational corrections, mine was taken using the Canon 40D. That was a 10 megapixel beast new in 2007, which offered what should charitably be described as a somewhat limited dynamic range, so I had to expose three bracketed shots, and combine them using HDR software. This was the result:

Original HDR image (Show Details)

The 40D was an APS-C format camera and at the time my best “normal” lens was the 17-85mm, equivalent of about 27mm at the wide end. This forced upon me a less than ideal high shooting position which generated a geometry with diverging verticals. Fortunately even in 2008 Photoshop could do a suitable geometric correction which at least straightened up the window and horizon:

Adjusted HDR Image (Show Details)

That’s not bad, but if you look carefully you can see that the image has a typical “early HDR” feel, with soft details, muddy colours, and noticeable coloured fringes along sharp edges courtesy of the 17-85mm’s significant chromatic aberration. Bob’s post therefore got me thinking about how my image might look with modern processing.

Fortunately I had kept the original RAW files, so it was the work of moments to open them up in Capture One. The latter immediately improves matters with its “HDR” shadow and highlight recovery, and allows me to work with just the central image of the bracketed set, so goodbye to composite HDR. Capture One also has correction for the lens’s chromatic aberrations, and allows me to apply the required geometric correction non-destructively. This is the output image:

Central image processed with Capture One (Show Details)

That’s a major step forward. However pixel peeping reveals that the image is still not as sharp as would be ideal, and despite being shot at ISO 200 has noticeable noise in the darker areas. There’s also the compositional issue of the power socket half visible behind my Pina Colada. Fortunately we can also pass the image through Topaz Photo AI. It volunteered sharpening, and I also applied its denoise technology. I then went into its “remove” function and highlighted the half socket. This is the result:

Central image reprocessed with 2026 software (Show Details)

I love the way this “pops”, and we’ve got rid of the compositional compromises. I’d be happy with this image if I took it today with a modern camera and much more expensive lens. Ideally I would actually have to hand a wider lens which would allow me to get the whole of the right shutter in shot, but I can live with it.

This is great news for cases like this, but it’s amazing news for cases where the old image can no longer be recreated, because the people or architecture have moved on. So if you have in your collection images of which you’re fond but which are technically limited, and you still have the RAW files, try reprocessing them. The results may be wonderful.

View featured image in Album
Posted in Photography, Thoughts on the World | Leave a comment

The Unhelpful Amp-hour

Omni-charger

I spent much of the last few years working on a large company’s Net Zero project, within which a significant element of my role was trying to educate people to understand electrical power and emissions calculations.

It was hard enough training those without much scientific background to understand power in Watts, energy in Watt-hours, and their various standard multiples in SI units. At least the relationship between power and energy is a straightforward one.

However once we started talking about energy stored in and used from batteries, we encountered a whole new level of complexity. That’s because contrary to common sense, battery drain and capacity are almost always quoted not in Watts and Watt-Hours, but in Amps and Amp-Hours. Amp-hours (Ah) are a measure of the current flowing through a circuit, multiplied by (or, strictly, integrated over) time. If a steady current of 1 Amp flows for 1 hour, that’s one Amp-hour. If that circuit consuming 1A is powered by a battery marked 10Ah, then you know the battery should be able to power it for 10 hours.

So far so good. If you know the current drain, and everything runs at the same voltage, you can do these calculations easily. Older car electrics which all work at 12V are a standard example. Then there’s a great sequence in Apollo 13 where Ken Mattingly (played by Gary Sinise) sits in the simulator and experiments with sequences to wake up the stricken craft to find one which fits the capacity and the maximum current of the usable fuel cells, using just an ammeter.

The Amp-hour model breaks down

Unfortunately it all rapidly breaks down in our modern world of multiple battery technologies, with varying loads and voltages.

Here’s a simple example. I have a powerbank battery, which is marked 20Ah. I want to use it on a trip to recharge my drone batteries. According to Google the DJI Mini 3 Pro uses batteries with a 2453 mAh (≈ 2.5Ah) capacity. How many batteries can I recharge, and how long should it take?

You might think that the answer to the number of batteries is 20/2.5 = 8. But it’s not that, or anything remotely like it. That’s because the quoted Amp-hour figure is "at a certain voltage". Change the voltage and you change the Amp-hour capacity.

If you’re not sure about this, think about two AAA cells each with about 1Ah capacity. Wire them up in series and you get a voltage of 3V, but only 1Ah can flow (the same current flows simultaneously through both). Wire them in parallel with a switch between them, the output voltage is 1.5V but you’ll get 1Ah out of each in turn, so 2Ah in total.

The problem with the powerbank and drone batteries is they have different native voltages, and 1Ah therefore represents a different amount of power in each case.

OK, so I’ve convinced you that we should think about power/energy rather than current. Well Power = Voltage * Current, so that should be easy, right?

Sadly not. Google helpfully confirms that "the DJI Mini 3 Pro uses 7.38V Li-ion Intelligent Flight Batteries, with a maximum charge voltage of 8.5V". So what voltage should we use? 7.38V, 8.5V, something in between? Fortunately a bit more Googling confirms that that the batteries have an energy capacity of 18.1 Wh, so working backwards we get a voltage of 18.1 / 2.453 = 7.38, so that’s the right one, but you could easily get it wrong.

It gets even more confusing with the powerbank. They don’t sell exactly the same model any longer, but the closest replacement supports output voltages of 4.5V, 5V, 9V and 12V. At which of these voltages is the Ah figure quoted? It turns out to be none of them. Hidden away in the 2pt text in pale grey on the bottom of the battery is the capacity in energy terms, 74Wh. Working backwards, that tells us that the Ah figure is quoted at 74 / 20 = 3.7V, which isn’t actually an output voltage. In this case we’re entirely dependent on the battery manufacturer providing a Wh figure. In fairness to Romoss their newer adverts all do, but they still use the almost meaningless Ah number as the "hero" figure. Many other manufacturers don’t bother – I couldn’t find a Wh figure for any of a random sample on Amazon.

Now we have two energy figures in Wh, we can finally answer my first question. The ratio between powerbank capacity and drone battery capacity is 74 / 18 = 4.1. It’s not actually going to be that good, because charging one battery from another is nowhere near 100% efficient, but 3 charges might be achievable, and that’s what I saw in practice.

The "how long should it take" question is a bit more tricky, as we don’t know exactly what voltage the two batteries will negotiate over USB-C. 9V is closest to the 7.4V of the DJI batteries, and Romoss also usefully state that the powerbank can deliver 2.2A at that voltage. A full charge is 18Wh. 9V x 2.2A is about 20Wh, but let’s assume about 75% efficiency like with the capacity figures. That suggests a full charge time of just over an hour, again matching my experience.

Another example

Once you’re alert to all this you find other examples everywhere. I have a number of Wifi-enabled water detectors in the house. I noticed that immediately after putting in freshly-charged Duracell rechargeable batteries, they report a battery level of ~67%, and the batteries don’t last as long as expected. Why?

The answer is that 1.5V does not equal 1.5V. Well it does, but what different manufacturers and technologies deliver as a "1.5V" battery varies widely.

Out with my trusty AVO meter, like me in its 7th decade, and still running on its original battery. Impressive. Let’s look at 3 AA cells in my battery drawer:

  • Duracell NiMH rechargeable battery. Output 1.3V unloaded. Stated capacity 2500mAh @1.2V, so energy capacity 2.5 * 1.2 = 3Wh
  • Varta Longlife alkaline battery. Output 1.5V. Published capacity 2750mAh, so energy capacity 1.5 * 2.75 = 4.1Wh
  • Hixon Li-ion battery. Output 1.5V. Stated capacity 3500mAh, so energy capacity 1.5 * 3.5 = 5.25Wh

Clearly the combination of low voltage and low capacity of the NiMH batteries explains their relatively short life between charges. Unfortunately the Li-ion batteries are dramatically more expensive and I’ve had a few reliability issues, otherwise they’d solve a lot of problems.

Is the Amp-hour completely useless?

Current is easy to measure. Almost as soon as we’d realised that current flowing through a coil produces a magnetic field, we added a magnet and a spring and created a meter which could measure current. Add a resistor, and you could measure voltage. Add some diodes and a capacitor and you could do the same tricks with alternating current.

Power is much more difficult to measure directly. In reality most of the time we measure current and (if we don’t know it) voltage and do the multiplication sum.

If you’re working with a sealed system then current and Amp-hour figures may be good enough. As I noted above, generations of electricians have operated that way. For a lot of purposes I don’t care what the voltage of my phone’s battery is: I can see the current drain over time with an app like AccuBattery, and I can work out how that matches the battery’s capacity in Ah. However as soon as you have to interface with external elements (e.g. "how many times can my powerbank charge my phone?") this breaks down.

Unfortunately the market is following tradition, in an environment where the tradition is no longer applicable. Why do powerbank manufacturers think it’s acceptable to quote an Amp-hour capacity tied to a mysterious internal voltage which isn’t even one of the options on the output? As a start, let’s try to quote stored energy figures in Wh as well, but in time they should become the primary figure, with the Ah consigned to a secondary role.

View featured image in Album
Posted in Sustainability, Thoughts on the World | Leave a comment

Review: A House of Dynamite

A House of Dynamite

Warning: contains spoilers.

I had been looking forward to Kathryn Bigelow’s new film for Netflix, A House of Dynamite. On the face of it this should be exactly our sort of film. Vantage Point, the 2008 film which shows an attempted assassination of an American President from multiple perspectives is a firm favourite. A House of Dynamite promised to be something not dissimilar in structure, with the twist of portraying the challenges of decision-making in the heat of an active nuclear strike.

The film promised to be story and character-focused and therefore ideally suited to Bigelow’s directorial style and the constraints of TV films with their limited budget for action, special effects and location. Unfortunately what results is a phenomenally boring plodding procedural in which the characters fail to behave in a professional manner as one hopes their real counterparts would do, and everything is left unresolved by a "non-ending".

Unlike Vantage Point which carries on past the shooting and explosion to complete the story, this film just stops, short of the missile’s impact. We never learn what happens next, or why. I understand that the writers and director were trying to focus on the challenges of making crucial decisions under extreme pressure and with limited information, but in reality those decisions would wait on at least confirming that a real bomb has exploded.

The story never proceeds past the first, lone missile. We don’t even know if it actually hits, or actually explodes when it does so. There is some discussion about the urgency of deciding on a retaliatory strategy, but given the uncertainties including who fired the missile and under what political or military orders one can’t help thinking that in reality one would wait until some of that uncertainty had been resolved. It’s shown that the US already has its planes and submarines ready to respond and waiting until more information was in place would not preclude a response if one were required.

There’s no real attempt to resolve the origin or purpose of the strike, which to my mind is critical to the response decision process. We’re told as a plot device that somehow US surveillance missed the actual launch, but surely it should be possible to trace the trajectory back? Can nothing be deduced from the flight path characteristics? Is there any useful signals or human intelligence?

To my mind the options include:

  1. Genuine first strike sanctioned by the regime of Russia, China or North Korea. This seems unlikely as there’s no obvious motive, and the characters don’t discuss any geopolitical context which might explain it. Also surely more than one missile would be launched, and there would surely be a more obvious attempt to disable the US ability to respond. Also destroying a major US city guarantees a massive response in the way a lesser or military target might not. Given the uncertainty about the origin, against whom do you respond? There’s literally no evidence for this option, and hence no justification for the "bomb everybody" response that the military characters are suggesting.
  2. Official but dummy strike to probe US capabilities, readiness and resolve. In this scenario the missile either destructs a few miles short of Chicago or makes a large but harmless splash in Lake Michigan. A lesser response is required, and you still need to work out against whom.
  3. Deliberate launch by terrorists or rogue elements acting independent of main regimes. We are shown the Russian foreign ministry desperately trying to say "it’s not us" and told China is saying something similar, which supports this option. Unlike in Crimson Tide or Hunter Killer there’s no background context such as a Russian insurgency to support this explanation. Again if this is the case it demands a more measured response, and you need to work out against whom.
  4. Accidental launch, due to either a human or technical failure. This is a horrible possibility. Yet again any response has to wait until you understand the outcome of the impact and the missile’s origin.

I don’t know how accurate a portrayal the film is of the process under which the president would make a decision about a response, but you can’t see even President Trump making a snap decision to simultaneously attack all America’s enemies and start WW3 without more information.

Some other plot devices are also not credible. The president is hurried onto Marine One, a large helicopter in perfect condition, but is accompanied only by the army officer with the nuclear football. What happens to the president’s assistants and advisers who were also at the basketball game and might be able to help analyse the situation?

The isolation of the president is enhanced by unrealistic comms issues. We’re shown one of the National Security Advisers joining the central call by video from a phone even as he’s running around Washington and going through security, but the president is restricted to a poor quality audio-only link from The Beast and Marine One. The vast White House machine is unable to patch calls together, so more junior characters have to take key calls and then relay to more senior characters third hand. Really?

A very surprising number of central characters behave in completely unprofessional ways, having breakdowns, focusing on personal considerations rather than their duties, even committing suicide. At one point Rebecca Ferguson’s character, who is in charge of the situation room, sends her deputy out to fetch personal mobile phones from their lockers. You think maybe she’s come up with a cunning solution to the problems with the White House phone system but no, she makes a personal call and he stands catatonic scrolling through pictures of his family. While it is certainly true that not everyone would live up to the expectations of their role, these people are chosen and trained to do so and you would hope more would get on with it.

This film is a complete waste of an opportunity and a strong cast. If I had paid to see it at the cinema I’d be wanting my money back. Fortunately it’s part of my Netflix subscription, which is currently otherwise good value, but it was still a frustrating waste of my time.

View featured image in Album
Posted in Reviews, Thoughts on the World | Leave a comment

Adjustable Architecture and Minimising Waste

Back to Normal
Camera: SONY DSC-RX100M7 | Date: 22-10-2025 15:14 | Resolution: 5322 x 3326 | ISO: 100 | Exp. bias: -0.7 EV | Exp. Time: 1/50s | Aperture: 5.0 | Focal Length: 12.3mm (~33.0mm)

Rescue, Don’t Replace

One of the things which attracted us to our house about 30 years ago was a great feature: what is known as a “Chinese Circle” in the courtyard end wall, which provides a view into, from and through the courtyard from both the house and the garden. We place sculptures so they are viewed through it, we light it at Christmas, it’s very much part of what makes our house.

Unfortunately a few years after we moved in, it became apparent that the original wall had not been built very strongly and was in some danger of collapse. Therefore about 25 years ago we had it rebuilt. We contracted a local builder who agreed a much stronger double-thickness structure, plus what we hoped were adequate, larger foundations.

While the wall itself was impressively strong, we’re on clay and over the years it became clear that with each cycle of wet and then hot weather the foundations were moving slightly. In recent years this accelerated, with the wall moving by several millimetres this summer and getting to the point where there was some risk of collapse. The wall itself was still stable and uncracked, it was just leaning into the garden, as a whole, by about 5°.

The Leaning Wall of Effingham (Show Details)

Knowing the wall was still stable I approached several subsidence specialists. They made it very clear they were not interested in such a small job, claiming that the simplest solution was to knock the wall down and rebuild it. We approached a reputable local builder who said very much the same thing: he was happy to provide a quote for rebuilding, at enormous cost, but when we pressed him for a quote to stabilise it he basically refused, by providing a quote effectively no different to the rebuild option.

Apart from the impact on our finances this just felt wrong and wasteful. Other things aside, a rebuild would require at least £1000 worth of new bricks, with the existing ones being disposed of as rubble. The wall was strong and undamaged apart from leaning. Even if it could not be fully righted, it would be acceptable to just stabilise and support it where it was. Why was no-one prepared to do that?

With another local builder Frances and I came up with the idea of creating two steel buttresses to stabilise the wall in place. We were quite keen on the option, but it became apparent that his steel fabricator was going to charge a fortune to make up the buttresses, and we’d pay a lot of money for an aesthetically questionable part solution.

Just as we were in danger of exhausting the local directories, we were introduced to friend of a friend Tomasz and his team of Polish builders. Initially he wanted to quote for a rebuild, but when pressed he agreed that it should be possible to jack the wall back nearly vertical, and then underpin it. The quote to do so wasn’t much more than half the rebuild option, he was able to start almost immediately, and we almost bit his hand off.

On the anointed day up turned a team of Polish chaps with shovels and a tiny digger, who proceeded to dig two deep trenches either side of the wall. I was a little afraid that it would collapse during this process but with strategically placed props and wooden supports it they managed to keep it all in place.

The first attempt at jacking used two steel props with contact points halfway up the wall. This quickly reduced the lean angle by about half, but we were concerned that the wall might crack at its base if the wall moved and the foundation didn’t. We were prepared to stabilise the wall at the new position, but the Polish guys went back to digging and created a new structure in which two smaller jacks could be used to twist the foundation itself.

The next challenge was finding the right jacks. They had one small hydraulic jack, pretty good, and a bunch of modern car jacks which were clearly not going to work. However I rummaged in the back of my garage and found a bottle jack rescued from an old Ford Transit in the 1970s which turned out to be exactly the right piece of equipment. Twenty minutes of careful jacking on the foundation twisted it with the wall intact, and we had a perfectly straight wall again.

My faith in the strength of the wall was fully vindicated – it didn’t crack or warp at all.

The rest of the process was straightforward albeit physically hard work, progressively digging by hand and pouring extended foundations which were wider and deeper than the old one.

We decided to cap off the new structure by building two brick buttresses on the garden side of the wall. While not strictly required because the wall is still uncracked and now vertical they enhance the look of it and make sure if there is any cracking right at the base of the wall it won’t compromise the solution. Tomasz procured 100 matching bricks, and I was given the task of coming up with a design for their use in the buttresses. My design used 98 of them. We also needed to cap off the top bricks, the dimples of which would otherwise collect water. I came up with the solution of creating porcelain tile caps, and then impressed the guys by getting out my own electric tile cutter, and making the required caps from a single yellow floor tile left over from our 2006 bathroom refit.

Brick Buttresses (Show Details)

Overall the process took an average of two men just over a week. The bill was under 2/3 of the cheapest rebuild quote. The excavated clay had to be removed from site, but otherwise there was zero waste, apart from two spare bricks!

The guys tidied up and disappeared, making good so well you’d never know they’d been there. The next day we watched an episode of “Grand Designs” in which the house was pre-fabricated 200 miles away from the plot, and moved as completed modules which were craned onto waiting foundations. Although the process was relatively painless, it was enormously expensive, and there were a few points where the prefabricated structure had to be hacked about with axes and chisels to accommodate unexpected service positions, which just felt wrong. Essentially the structure didn’t allow adjustment.

Good architecture should be accessible and adjustable. I’ve always believed in this for both the software I have designed in my professional career, and the hardware solutions I have developed for our living space. For example, I make sure that pipes and wires run in accessible spaces, and allow for change. However I’ve usually accepted that this might not be possible with the lower levels of physical architecture, the “Structure” layer of the Frank Duffy / Stewart Brand model.

Now I’m not so sure. We stuck to our guns, and we adjusted a brick wall!

 

With a big dziękuję to Tomasz, Rafal and Artur.

View featured image in Album
Posted in Agile & Architecture, Thoughts on the World | Leave a comment

Acceptable Update Strategies

Please wait
Resolution: 2848 x 1600

Excellent

Example: Microsoft Visual Studio. You finish your work, and when you exit from Visual Studio, it prompts you with “Updates are available, would you like to install them now?”. There are Yes and Cancel (= defer to next time) options.

Good

Examples: Idea IntelliJ, Topaz Photo. While you are working the program checks for updates in the background. If they are available it shows a subtle notification which does not get in the way of your work. When convenient, you click on the notification to start the installation process.

Also Good

Example: Windows Update. A background process detects available upgrades which are silently installed while you work. If they can be completed without interrupting usage, that’s done. If not they are deferred until either you choose to reboot the machine, or it can be done automatically when not in use.

Acceptable, Most of the Time

Example: Microsoft Office 365. A background process detects available upgrades, which are silently installed while you work. Occasionally there’s an annoying notification that Office needs to be closed down for a short period to complete installation, but you have the option to defer.

Wholly Unacceptable

Examples: most meeting and messaging apps. You go into the software just before your meeting. It looks for updates and if they are available starts to install them. There is no option to defer.

Why, just why?

View featured image in Album
Posted in Thoughts on the World | Leave a comment

A Bit of Variety

Female detectives

One of the great things about watching a lot of cop shows on television is the endless variety of mechanisms used to set up key characters. Recently we’ve had…

The Island: female detective born on Harris returns there after several years serving with the police in central Scotland.

Shetland reboot: female detective born on Shetland returns there after several years serving with the Met in London.

Return to Paradise: female detective born in a small town in Australia returns there after several years serving with the Met in London.

Get Millie Black: female detective born in Jamaica returns there after several years serving with the Met in London.

The One That Got Away: female detective born in Pembroke returns there after several years serving with the police in Cardiff.

Borderline: female detective born in a small town in Ireland returns there after several years serving with the Gardi in Dublin. OK, she then has to deal with cross-border interactions reminiscent of The Bridge working with a male detective from the PSNI, but still.

Thank god for The Turkish Detective! Some variety at last!

The Turkish Detective: male(!) detective born in Istanbul returns there after several years serving with the Met in London. Oh well…

View featured image in Album
Posted in Thoughts on the World | Leave a comment

(In)Correct Tripod Technique!

Who Needs a Drone?

With Apologies to My Photography Tutors

First, I’d like to apologise to all the authors, tutors, mentors and tour leaders who have tried to instil in me “correct” tripod technique. As they say, it’s not you, it’s me.I don’t particularly have a philosophical objection to the three legged beast, and you will catch me using one when the camera has genuinely got to stay in one place for a long time, for example videoing a wedding speech, or taking a time-lapse photo sequence. However I do prefer a more free-wheeling approach, and my discovery that the latest Panasonic Dual Image Stabilisation allows me to hand-hold a wide-angle shot for 2-3s, or an extreme telephoto for 1/5s, has only encouraged me to eschew support when not essential.I do actually own three of the things, but the large and relatively heavy Manfrotto has only been used on one single photo shoot. However, coupled with the laser level it has proven invaluable as a surveying tool for our kitchen refit and preparing for my new workshop. My medium-sized Gitzo is a solid, professional but lightweight device, my default for a “serious” tripod. However I suspect I’d scare the purists by pointing out the pistol-grip Manfrotto head – none of this fancy gear rubbish!

And then there’s the Frankentripod ™. (It’s great to have one search term which gets your website at the top of all the search engines, even if you have to make up a word to do it!) It’s the tripod I carry when I don’t want to carry a tripod, but feel I should, to show willing, or “just in case”. It does the job, most of the time, at very little cost in terms of luggage space or weight. I say “most of the time” because there are limitations – in Cinque Terre, for example, it wasn’t quite tall enough to clear some of the safety fences around the harbour in Manarola, but I found work-arounds.

The Frankentripod also turns out to be the ideal device for my new tripod technique, which I call “why get a drone when you have a tripod?” The technique is simple:

  1. Extend tripod to full length, legs together
  2. Put camera on tripod
  3. Flip screen out so you can see it from below
  4. Put camera on self-timer
  5. Start self-timer
  6. Raise camera above your head, to clear obstacles such as vegetation, tall people etc.
  7. Compose your shot by tilting the tripod as required
  8. Wait for 0s
  9. Bingo!

In the right situation this can be a game changer. For example, our views of Montisi from Castelmuzio were somewhat constrained by the fact that whoever was responsible for trimming the vegetation at the viewpoint was somewhat derelict in their duty. From head height the best wide shot was this, with some rather intrusive branches:

Montisi from Castelmuzio (Show Details)

However, the Frankentripod method got my camera up above the errant foliage, and produced this:

Montisi from Castelmuzio (Show Details)

The method does have some pre-requisites, and a couple of trust issues:

  1. Lightweight camera and tripod. This is not going to work with a Canon 1DS, 600mm lens and Uranium tripod, unless you take my instruction in point 5 very seriously!
  2. A flip-out or flip-down screen bright enough to be viewed from the other end of the tripod
  3. Image stabilisation, very good image stabilisation
  4. Trust that your camera is firmly attached. Otherwise this could become somewhat dangerous and expensive, although a long strap (shorter than yourself, obviously), tied to the tripod head, and a hard hat might provide some mitigation
  5. A steady hand and strong shoulders, otherwise ditto. Pump iron!
  6. A bit of skill to compose and then hold the composition in 10s, or whatever the timer is on your camera.

Or I could just buy a drone, Darling Wife!

View featured image in Album
Posted in Italy 2024, Photography, Thoughts on the World, Travel | Leave a comment

Trippin’ AI

Trippin' AI
Resolution: 1792 x 1024

Just how wrong can an AI get it?

As part of my effort to profile the power consumption of GenAI, I decided to try and summarise one of my travel blogs using ChatGPT and the other big public models, plus a series of smaller models which can be hosted locally on my Alienware workstation. What’s interesting is just how well, and badly, the different AIs handled the task.

The instruction in every case was the same: “Summarise the blog at https://www.andrewj.com/blog/2024/three-kirks-and-a-kirkufell/”. You might want to read the blog yourselves before reading further: it’s neither long nor complicated, but it defeated most of the contenders.

I tried to avoid writing “WTF”, but I was sorely tempted. Some of these are hallucinating like they’ve just taken LSD…

The Attempts!

ChatGPT

The blog post “Three Kirks and a Kirkufell!” recounts a photography trip around Iceland’s Snæfellsnes Peninsula. The author, despite dealing with a cold and Iceland’s challenging weather, visits and photographs three distinctive churches (“kirks”)—Búðir, Stóra-Hella, and Enni—before capturing the iconic Kirkjufell mountain. The journey includes dramatic landscapes, with highlights like waves crashing and snow-covered scenes. The narrative reflects the difficulties and rewards of winter travel in Iceland.

Score: 10/10. Excellent, couldn’t do better myself.

Gemini (Bard)

This is an article about a photographer’s trip to Iceland. It discusses his experiences photographing three churches and Kirkufell. He had some difficulty finding pharmacies with the medications he needed. He also had some trouble finding Kirkufell because of the weather. He eventually found it and took many photographs.

Score: 6/10. OK, effectively correct apart from the bit about not being able to find Kirkufell, but lacking the nicely condensed detail of the ChatGPT version.

CoPilot Chat

The blog post titled “Three Kirks and a Kirkufell!” by Andrew Johnston describes a travel experience in Iceland. The author recounts a cold, snowy day driving along the north of the Snaefellsnes peninsula. The journey included an attempt to visit Kirkufell, a notable mountain, but the stop was abandoned due to poor visibility from the snow. The post highlights the challenges and beauty of traveling in Iceland during winter.

Score: 4/10. Just adequate. Some factual inaccuracies and poor detail, but at least still rooted in reality.

Wizard V1.2

The article “Three Kirks and a Kirkjufell” is about an adventure trip to Iceland by the author, Andrew J., who visited three churches (Kirks) in Reykjavik. The first church was Hallgrimskirkja, which has a unique architecture that resembles a rocket ship taking off into space. The second one was the Lutheran Church of Holmenkollen, known for its beautiful stained glass windows and traditional Norwegian wood carvings. The third Kirk was the Gamla Stan Church in Stockholm, Sweden, an old church with a rich history dating back to the 13th century.

Andrew J. also visited Kirkjufell, a mountain near Reykjavik that is known for its unique triangular shape and appeared in the popular TV series “Game of Thrones”. He enjoyed hiking around the base of the mountain and admiring its beauty from different angles. The author concludes by sharing his love for exploring new places and experiencing their cultures, architecture, and natural wonders.

Score: 2/10 (feeling generous). Starts OK, but rapidly descends into just making stuff up. How I’m supposed to have teleported to Stockholm while in Reykjavik is not explained.

Llama 3B Instruct

The blog post “Three Kirks and a Kirkufell” by Andrew J. is an entertaining account of his experience hiking in Iceland’s stunning landscapes. Here’s a summary:

Andrew, along with friends Sarah and Mike, embarked on a 7-day adventure to explore three iconic KIRKs (churches) and the famous Kirkjufell mountain in western Iceland. The trio aimed to visit the churches at Snæfellsnes Peninsula, Borgarnes, and Reykjavik. Their journey began with a scenic drive along the Golden Circle route, where they marveled at Þingvellir National Park’s unique geological features. They then visited the picturesque church in Borgarnes, which dates back to 1615. The next day, Andrew and friends explored Snæfellsnes Peninsula, admiring its breathtaking landscapes, including black sand beaches, glaciers, and volcanic mountains. They stopped at a charming church in Arnarstapi village, built in 1869. Their final destination was Kirkjufell mountain, made famous by Game of Thrones as the “Arrowhead Mountain.” The group hiked to the base of the iconic peak, taking in its majestic views and snapping photos. Throughout their adventure, Andrew shares witty anecdotes about his friends’ quirks (or lack thereof), amusing observations on Icelandic culture, and stunning visuals showcasing Iceland’s natural beauty.

Score: 2/10. The start and end are OK and at least the context is correct, but the bit in the middle is complete fantasy.

Llama 3

The title is “Three Kirks and a Kirkufell”, and it’s a book review of “The Fellowship of the Ring” by J.R.R. Tolkien. Here’s a summary:
The author, Andrew J., reviews his re-read of “The Fellowship of the Ring”, the first book in The Lord of the Rings trilogy. He reflects on how he appreciated certain aspects more this time around, such as the character development and world-building. Andrew praises Tolkien’s writing style, saying it’s both engaging and immersive. He also discusses the themes of the book, including friendship, loyalty, and the struggle between good and evil. The reviewer notes that the pacing is well-balanced, with a mix of action, humor, and quiet moments. Andrew concludes that “The Fellowship of the Ring” remains an excellent fantasy novel even after multiple readings, and he’s excited to continue with the rest of the trilogy.

Score: 0/10. This is a complete mystery, especially given that a slightly different version of Llama at least got the context right. How on earth did they get to “Lord of the Rings”? Ironically I could understand “Journey to the Centre of the Earth” because at least that starts at Snaefells…

Mistral

The blog post, “Three Kirks and a Kirkufell,” by Andrew J., discusses the author’s experience visiting Iceland and exploring various sites related to Star Trek. The three Kirks mentioned in the title refer to Captain James T. Kirk from the popular science fiction series, who has been immortalized in different ways throughout Iceland. These include a statue of him at Keflavik International Airport, another one near the Blue Lagoon, and an actual mountain called Kirkjufell (Kirk’s Mountain). The author also shares his thoughts on the unique landscapes and natural beauty found in Iceland, which inspired the creators of Star Trek to use it as a backdrop for their show.

Score: 0/10. Complete balderdash. This model obviously doesn’t recognise the work “kirk” despite the fact that it’s a regular part of Scottish English, with exactly the same spelling and meaning as in Icelandic. On a positive note at least we’re back in Iceland from Middle Earth…

GPT4All Falcon

The blog post “Three Kirks and a Kirkufell” by Andrew J. is about his experience hiking in Iceland, specifically the three peaks of Kirkjufell, Kirkjuvatn, and Kirkjufell. He describes the challenges he faced, such as navigating through thick fog and snow, and how he overcame them with determination and perseverance. He also shares some tips for other hikers who may want to attempt this hike. Overall, the blog post is a personal account of Andrew J.’s experience in Iceland and his love for hiking.

Score: 1/10. Poor. I don’t mention hiking, and by my count the summary lists two mountains not three. Again no recognition of “kirk”.

Ghost 7B

Summary:

The blog post discusses three different churches: Kirk, Kirke and Kirche. Each church has its own unique characteristics and history. The author explains that the word “church” itself comes from the Old English word “cyrc,” which means “circle.” This is reflected in the shape of many early Christian churches, which were often circular or semi-circular.

Key Points:

o Kirk: A Scottish church type characterized by its low, round roof and lack of windows. It was originally designed to be a place for worshipping God outdoors.

o Kirke: A Norwegian church type that is similar to the Kirk in Scotland. However, it has a slightly taller roof and may have some small windows.

o Kirche: The German word for “church.” It refers to any Christian church, regardless of its size or style.

o Kirkufell: A term used by some Scandinavian countries to refer to the Church of England. It means “the church in the field” and is a reference to the fact that many early churches were built on agricultural land.

Conclusion:

The blog post provides an interesting overview of three different types of churches: Kirk, Kirke and Kirche. Each type has its own unique characteristics and history, and they all play an important role in the religious life of their respective countries.

Score: 0/10. We’re off on a tangent again, nothing to do with the source, and as far as I can see they’ve just completely made up some of the details.

MPT Chat

The article discusses how Andrew Kirk has been inspired by three different individuals in his life – a friend, an uncle and father figure who taught him about business ethics; the Dalai Lama for teaching him mindfulness techniques to deal with stress management at work. He also shares some of these lessons learned from them that have helped shape their own approach towards leadership which is centered around empathy rather than control or manipulation.

Score: -5/10. Nope, I haven’t got a clue how they got this from what I wrote.

Conclusions

The three online models (ChatGPT, Gemini and CoPilot Chat) all worked, although there is a very marked difference in quality and detail of output. The self-hosted models all failed, some more miserably than others…

In a large number of activities, better training and access to better resources both play a part in the success rate, and it appears that GenAI is no exception. Models you can host on your own system (even with a moderately high-spec GPU like my NVidia 3060 with 12GB RAM) are just pale shadows of the “very large language models” such as ChatGPT. They have neither the resources to hold significant knowledge, nor the training across a wide enough knowledge base. That doesn’t preclude effective use on a tightly-focused specialist use case, but their generalist capability is going to be very limited.

Beware, but also be afraid, be very afraid.

View featured image in Album
Posted in Thoughts on the World | Leave a comment

An AI Scares Itself, and Me

Dystopian Landscape
Resolution: 1792 x 1024

Just how bleak can an AI’s world view become?

One of my clients asked me to write an article on the environmental impact of generative AI. Like a lot of large corporations they are starting to embrace GenAI, but they are also a leader in sustainability, aiming to be one of the first major financial organisations to Net Zero, and so they are rightly worried about the potential adverse effect of large-scale use of GenAI. I’ve already been doing work with them to measure and to better understand the power consumed by GenAI, so the article is an obvious next step.

I’ve already written several articles for their internal Net Zero blog, and to add a bit of interest to each I’ve asked ChatGPT to generate a short essay on the topic, in the style of a well-known author. This appears to be a task at which ChatGPT excels, and I saw no reason not to do the same again.

God, That’s Bleak…

I like the way ChatGPT mimics the very direct writing style of Hunter S Thompson, and I like to think that if he was still with us, Thompson would be working to expose shortcomings on all sides of the climate change debate. The prompt to ChatGPT was therefore a simple one: “Write a short essay about the risks of AI with a focus on the environmental impact, in the style of Hunter S Thompson”. This was the result:

A Savage Journey Into the Heart of Techno-Doom: AI, the Environment, and the Goddamn Reckoning


We live in a time of extraordinary madness, where artificial intelligence (AI) has become the darling of our cursed, frenetic age. It’s a glittering gem, a tool of infinite promise—yet lurking beneath this polished veneer lies the venomous bite of catastrophe. If you think AI is just another shiny toy for the tech elites to fiddle with, you’re in for a rude awakening, my friend. The real story here is not just the dystopian nightmares of machines outsmarting us; it’s the environmental havoc that this digital beast is ready to unleash on an already bleeding planet.

Let’s be blunt—AI is a power-hungry, resource-guzzling monster. It doesn’t just run on hopes and dreams; it feasts on data and electricity, and the price is paid in the blood of the earth. Picture it: vast data centers sprawling like cancer across the globe, consuming as much energy as entire nations. These steel-and-silicon behemoths are the new factories of the digital age, but instead of churning out cars or canned goods, they spit out algorithms and predictions. And just like the smoke-belching mills of the industrial revolution, they’re leaving a trail of destruction in their wake.

But who cares, right? So what if AI needs a few gigawatts here and there? We’re advancing, after all! The future is here, and it’s smarter than ever—never mind that the goddamn planet is burning. That’s just collateral damage, a minor inconvenience on the road to progress. This is the kind of twisted logic that’s driving us headlong into the abyss. The tech lords promise us a utopia of endless convenience, while quietly brushing the environmental costs under the rug. But you can’t hide the truth forever.

Let’s get real about the energy. Training AI models, particularly those gargantuan neural networks that gobble up every scrap of data they can find, is like fueling a fleet of private jets with coal. The carbon footprint of these operations is staggering. Every keystroke, every click, every goddamn chatbot response comes at the cost of churning out more CO2. And let’s not forget the hardware—the servers, the cooling systems, the endless upgrades. They all need minerals mined from the earth, ripped from the ground in a frenzy that leaves behind a scarred landscape and poisoned water.

And then there’s the e-waste. This relentless march of technology demands constant upgrades, and when those shiny new devices become obsolete, they’re tossed aside like yesterday’s trash. But this trash doesn’t just disappear. It festers in landfills, leaching toxins into the soil, contaminating water supplies, and creating a slow-motion disaster that will haunt us for generations.

But here’s the kicker: as AI becomes more “efficient,” as it gets better at predicting, automating, and optimizing every facet of human existence, it only exacerbates the problem. The smarter it gets, the more data it needs, the more servers it demands, the more energy it consumes. It’s a feedback loop of doom, an ouroboros of technological advancement eating its own tail, and we’re all along for the ride, whether we like it or not.

So what’s the answer? Do we unplug the machines and head for the hills? Not likely. The AI genie is out of the bottle, and there’s no stuffing it back in. But we need to start facing the reality of this techno-madness before it’s too late. This isn’t just about smarter machines or even smarter humans. It’s about survival. Our planet is at stake, and if we don’t get our heads out of the silicon-scented clouds, we’re going to find ourselves deep in the muck, wondering where it all went wrong.

In the end, AI might just be the ultimate test of human intelligence—a test of whether we can outsmart our own creations before they outstrip our ability to live on this Earth. If we fail, the machines won’t need to destroy us. We’ll do it ourselves.

 

Bleak, or what?

And how do I teach myself to write like that without an AI?

Cheerful, but Profound

I liked it, but unsurprisingly my clients were not that keen, especially as there’s a current internal campaign about the benefits of AI. So I tried again. Let’s not do bleak, let’s do cheerful. As a child of the 1960s I struggle to think of anything more cheerful and unthreatening than Winnie The Pooh, so I went back to ChatGPT with a second prompt: “Same again, please, but in the style of Winnie the Pooh”. Usually the “same again” instruction or similar resurfaces the same concepts, just wrapped in different language, but this time ChatGPT did something quite profound…

A Thoughtful Little Reflection on AI and the World Around Us

One fine day in the Hundred Acre Wood, while sitting on a sunny spot, Pooh Bear began to think—very hard, which is not something he did often, mind you, but this was a special kind of thinking. You see, there was talk in the Wood about something called Artificial Intelligence, or AI, which sounded very important and rather clever. So, Pooh decided he ought to try to understand it, especially since it seemed to be making quite a fuss in the world beyond the trees.

Now, Pooh was a bear of very little brain, as we know, but he had a great deal of heart. And when he thought about this AI, he began to wonder not just about what it could do, but about what it might mean for the world—especially for the bees, the flowers, and all the other important things that made life in the Hundred Acre Wood so delightful.

“AI,” said Pooh to himself, “must be a very busy sort of thing, always thinking and learning. But if it’s thinking and learning all the time, it must need a lot of honey to keep it going!” But of course, AI didn’t eat honey. It ate something else—electricity, and lots of it. And that gave Pooh pause. He remembered how much work it took to make a pot of honey, and he wondered how much work it must take to feed all that electricity to AI.

Pooh scratched his head, thinking of the bees buzzing in the flowers, gathering nectar bit by bit. “If AI needs so much energy,” he thought, “where does it all come from? And what happens to the Wood if we use too much?”

You see, Pooh had heard that the world was getting warmer, a little more each year, and that too much energy-using could be a rather bad thing. The flowers might not bloom as they should, the rivers might run dry, and the trees might stop growing so tall and green. And that would be a very sad thing, indeed. For without the trees, where would Owl perch? Without the flowers, where would the bees find their nectar? And without the rivers, where would Piglet splash on a hot summer day?

It seemed to Pooh that while AI was very clever, it might not always be very wise. For if it wasn’t careful, it could use up too much of the things that the world needed to stay happy and healthy. It was a bit like eating too much honey in one go—not a bad thing in itself, but too much of it might leave you with a tummy ache, or worse, no honey left for tomorrow.

Pooh also thought about all the things that AI needed to be made. It wasn’t just the electricity; it was all the metal and other bits that had to be dug up from the ground, and then made into machines that could think and learn. And when those machines got old, they couldn’t just be tossed aside like a worn-out honey pot. No, they had to be dealt with carefully, or else they might end up in a big heap, making a mess of the beautiful Wood.

So, Pooh decided that while AI might be a very useful and clever thing, it needed to be used with a bit of care, the same way one should be careful not to eat all the honey at once. He thought that maybe, just maybe, if everyone thought a little more about how to use AI wisely, and didn’t let it gobble up too much of the world’s energy and resources, then perhaps the Wood could stay green and lovely for a long, long time.

And with that, Pooh decided it was time for a little smackerel of something—because even a bear of very little brain knows that sometimes, a small break is the best way to keep thinking about big, important things. So off he went, humming a little tune and hoping that the clever people working on AI would remember to take care of the world around them, just as he took care of his honey pots.

 

There are some quite deep messages in here. I particularly like “It was a bit like eating too much honey in one go—not a bad thing in itself, but too much of it might leave you with a tummy ache, or worse, no honey left for tomorrow.”

What’s interesting is the different focus of the two pieces: the dominant message in the “Hunter S Thompson” one is very much “our rush for more technology is destroying us”, whereas in the “Pooh” one it’s “AI is a good thing, but not if we use it too much”, which is much more balanced.

Sadly I haven’t yet persuaded my clients to use either version, but at least I get to share them with you.

View featured image in Album
Posted in Agile & Architecture, Sustainability, Thoughts on the World | Leave a comment

Are British Airways a Bus Company?

'Nuff said

Are British Airways an airline or a bus company? You’d hope the answer was evident from the name, but I’m beginning to have my doubts. I’ve just done an analysis of the flights I’ve taken with them from Heathrow since 2018:

Morocco 2018. This was so ridiculous it’s laughable, but with a dark shade because it was really quite a bad health and safety failure. We sat at the gate for the outgoing flight and they called us forward by "group number". Instead of going down the ramp to an aeroplane, we had to go down several flights of steps and outside, where a number of buses were waiting. We were directly randomly to buses with no attempt to keep the boarding groups even roughly together. After a long ride we arrived at the plane in the middle of a field, which had two sets of steps set up. One bus went to the front, another to the back. This resulted in people with seats at the back boarding from the front, while some with seats at the back were boarding from the back. "Punch up" doesn’t quite cover it, the average bar fight in a film is better organised. Fortunately we managed to sort ourselves out, but literally "nil points" to the useless BA organisation, and thank the stars that no-one suddenly needed medical attention or worse.

We had two short-haul BA flights in 2019, to Belfast and Copenhagen. I can’t remember either involving a bus, but maybe I’m blanking it out.

Patagonia 2023. After a 13 hour flight from Argentina the plane landed in a field, and was met by a bus. The ride back to the terminal took so long that I noted on my blog that I wasn’t sure whether we’d actually landed at Heathrow, or Northolt!

Belfast 2023. Our scheduled flight was cancelled, and we got bumped to one next morning. Having waited on the tarmac at Belfast for about an hour we got underway, only to land at Heathrow and sit on the tarmac again. Eventually we docked at a gate, and went up the ramp, to be directed down the stairs and out of the building onto a bus. This then drove us round from the "international" side of T5 to the domestic one, a trip which took about 20 mins, despite the fact you could probably walk it in about the same time.

Las Vegas 2023. After a 10 hour flight we parked in a field, to be discharged onto a bus. The ride to the terminal wasn’t quite so interminable as some of the others, but long enough.

4/6, maybe worse.

A few years ago a comedian, I think it might have been Michael Macintyre, lampooned as peculiarly British the phenomenon of the replacement bus service, as seen through the eyes of a foreign visitor:

"But I have ticket for train?"

"Get on the bus!"

I see that BA are honouring the "British" in their name by simply extending this pattern.

View featured image in Album
Posted in Thoughts on the World, Travel | Leave a comment